The proteome can be viewed as constellations of interacting protein modules which are organized into organelles, molecular machines, and signal transduction networks. To date, the interaction partners for only a small fraction of expressed proteins are known, yet protein-protein associations are a major driving factor in understanding protein function and activity. Here we propose to integrate emerging mass spectrometry-based technologies to create a high-quality atlas describing the physical organization and connectivity of the mammalian proteome housed in widely used human cell systems. Using affinity-chromatography with mass spectrometry (AP-MS) techniques, three aims are proposed. 1) Independently express, purify, and catalog protein interactions for the majority of genes expressed in an HEK293 cell. This will include ~10,000 genes as a single, high-throughput and high-quality pass and will also include biological triplicate experiments for a selected set of highly interacting proteins (~3,000) for purposes of validation. 2) Both during and at the conclusion of aim 1, computational and informatics analysis will be performed for all proteins investigated. This analysis will be used to determine putative biological functions and place these proteins within a biological framework based on known activities for interacting proteins. Furthermore, at the conclusion of aim 1, all forms of the data (from raw MS files to fina informatics analysis) will be made accessible via a custom web-based interface and also deposited at NCBI for additional dissemination. 3) Upon the completion of aims 1 and 2 (expected to be done within the first 2 years of the project), ~2,000 proteins will be selected based on their network connectivity and putative biological function as baits for AP-MS experiments using six different cell lines. This aim will serve to both validate global maps and provide a unique window into cell- and/or tissue-specific protein complexes. In total, this work provides the launching point for global mammalian interactions studies, generating an important resource containing comprehensive and unbiased sets of physical interactions.